Seen as Green? Assessing the Salience and Greenness of Environmentally Friendly Packaging Cues

ABSTRACT Environmentally friendly products become increasingly popular. To be considered for purchase, they must be (1) sufficiently salient and (2) perceived as environmentally friendly (“green”). The current paper investigates whether implicit (packaging material) and explicit (eco-labeling) packaging cues can facilitate such salience and greenness perceptions. We conducted a mixed-method, within-subjects experiment using self-report measures to assess perceived greenness, and mouse tracking as an exploratory measurement method to assess salience. Finally, we explored whether green consumerism, as a personal value, moderates these effects. Results confirm that both implicit and explicit cues that signalize environmental friendliness positively influence perceived salience and greenness. Furthermore, we find that implicit and explicit cues seem mostly independent in affecting salience and greenness perceptions and that green consumerism is no prerequisite for most of these packaging effects to emerge. As such, this research provides important new theoretical and practical insights on the role of packaging in food marketing.


Introduction
Although consumers world-wide have become increasingly aware of their environmental impact, their excessive consumption patterns contribute to current ecological challenges (Stolz et al., 2013;Yoon-Na Cho, 2015). Especially food consumption has a big impact on the environment (Poore & Nemecek, 2018). Hence, one of the solutions to reduce the negative impact of food consumption on the environment is stimulating consumers to purchase less products, and if they do, purchase more environmentally friendly products. Environmentally friendly food products can be defined as products "that respond to basic needs and bring a better quality of life, while minimizing the use of natural resources, toxic materials and emissions of waste and pollutants over the life cycle, so as not to jeopardize the needs of future generations" (Oslo Roundtable on Sustainable Production and Consumption, 1994), which can be the reduction of meat consumption, the consumption of seasonal, local, and in some instances organically produced foods (Vermeir et al., 2020), and the consumption of products in a packaging with a lower environmental impact (Granato et al., 2022). In line with this, products in today's supermarkets are becoming increasingly environmentally friendly, meaning that they harm the environment as little as possible (Steg & Vlek, 2009). Although consumers value these environmentally friendly products (Rokka & Uusitalo, 2008) they do not always buy them (Alwitt & Pitts, 1996;Bech-Larsen, 1996;Jerzyk, 2016;Moisander, 2007;Thøgersen, 1999Thøgersen, , 2004Thøgersen & Ölander, 2003;Uusitalo, , 1990) and they remain the minority (Rex & Baumann, 2007;Van Doorn & Verhoef, 2011). To provide insights on how to stimulate environmentally friendly food purchases, we focus here on two important requirements in order for consumers to actually purchase environmentally friendly products: the product must be (1) sufficiently salient to be noticed, and (2) perceived as environmentally friendly ("green").
Previous research has shown that an important prerequisite for products being considered and actually chosen is being salient, in other words: they must be noticed (Krajbich et al., 2010;Van der Laan et al., 2015, 2017. In this light, it is worthwhile to consider that food is usually packaged and that product packaging does not only function as storage, but also as a communication vehicle that may attract attention (Ahmad et al., 2012;Bech-Larsen, 1996). Since different products are constantly competing for the consumers' limited attention, it is crucial to determine which food packaging characteristics are perceived as salient and have the potential to attract this attention .
Furthermore, it is possible to use the package to infer product characteristics (Van der Laan et al., 2012). Recent research has for example shown that consumers perceive food packaged in sustainable packaging as more satiating (Donato et al., 2021), or natural (Marozzo et al., 2020) than food packaged in non-sustainable packaging. People might also infer from the product package whether a product itself is environmentally friendly. Thus, for environmentally friendly products to be considered, packaging cues must not only be salient, but they should also evoke perceptions of environmentally friendliness ("greenness"; Magnier & Schoormans, 2015). Although there is quite some literature on the effects of different types of environmentally friendly cues in advertising on product perceptions (Hartmann & Apaolaza-Ibáñez, 2009;Schmuck et al., 2018) and the literature on the effects of packaging cues is steadily growing (Binninger, 2017;McGuicken & Palomo-Vélez, 2021;Popovic et al., 2019;Vila & Ampuero, 2007), findings do not yet shed light on whether and how different types of (implicit vs explicit) packaging cues might affect the perceived salience and environmental friendliness of those food packages.
A recent literature review revealed that consumers primarily rely on what is defined as implicit packaging cues (i.e., packaging material) and explicit packaging cues (i.e., eco-labels) to identify environmentally friendly food packaging (Granato et al., 2022;Ketelsen et al., 2020), and a recent cross-sectional study suggests that people may rely on these two packaging cues for determining whether a package is environmentally friendly (Herbes et al., 2020). Hence, implicit environmentally friendly cues in this study are packaging cues such as shape or material that consumers may associate with certain product attributes, but that not explicitly make claims about these attributes (e.g., a cardboard-look packaging might have a more sustainable 'look and feel' than a bright white packaging but is not necessarily more environmentally friendly; cf., Granato et al., 2022). Explicit cues, in contrast, directly express food product attributes (e.g., an eco-label, providing a 'proof' of environmental friendliness, such as organic or palm-oil free label; cf., Granato et al., 2022). However, it is not yet fully known whether implicit versus explicit packaging cues affect the perceived salience and perceived environmental friendliness of the food product itself, and whether these cues might interact. To address this research gap, the current research aims to explore the interplay of implicit and explicit packaging cues on consumers' product perceptions. Our overall research question for this study is: To what extent do different food packaging cues (i.e., implicit and explicit) influence salience and perceptions of environmental friendliness ("greenness") of food products?
Additionally, we will study whether this impact of packaging cues on salience and perceived environmental friendliness is influenced by the consumer characteristic 'green consumerism' (Sparks & Shepherd, 1992). As green consumers are more concerned about their environmental impact (Schuitema & De Goot, 2015), they are expected to be more sensitive to environmentally friendly cues and are more likely to recognize products as being environmentally friendly (Chen & Chiu, 2016;Magnier & Schoormans, 2015;Martinho et al., 2015).
The contribution of the current study will be threefold. Firstly, by testing the effectiveness of implicit and explicit environmentally friendly cues in an experimental setting, we cannot only quantify and replicate previous findings on the effectiveness of implicit and explicit environmentally friendly cues but, importantly, also discern which of these types of cues is most influential in affecting the salience and perceived environmental friendliness of the product. Secondly, a mouse tracking method will be used to complement the commonly used self-report measures. This allows us to explore whether this less subjective paradigm can reveal the dynamics that underly consumers' decisions in response to different food packaging cues. Thirdly, the findings can contribute to a more sustainable society by providing important practical insights on what (combination of) packaging cues could be best used to make sure that environmentally friendly products are salient and perceived as environmentally friendly. We can hereby provide guidelines and suggestions on what type of product packaging (i.e., choice of packaging material, eco-label, or both) most saliently and successfully communicates environmentally friendliness. This knowledge will enhance the chance of consumers choosing environmentally friendly products and thereby inflict less environmental harm within their everyday consumption choices.

Environmentally friendly packaging cues
While food shopping, consumers often make use of packaging cues to assess products, because they are not able to assess the product directly (Olson & Jacoby, 1972). Packaging can influence perceptions of products, especially of low involvement, nondurable products (Underwood, 2003) and thus plays an important role in food marketing as a communication vehicle (Behe et al., 2014). An extensive body of literature has focused on implicit (e.g., packaging design) and explicit (e.g., claims or labels) packaging cues and showed that they can influence consumer perception of food products, including product quality (Van Ooijen et al., 2017), taste (Becker et al., 2011;Van Rompay et al., 2014) or healthiness (Van Ooijen et al., 2016). We argue that packaging as a vehicle can also communicate environmental friendliness of a food product. Hence, we focus on the two packaging cues that are possibly used by consumers to determine whether a product is environmentally friendly: packaging material (which may serve as an implicit cue from which consumers derive environmentally friendly product attributes) and eco-labeling (which may serve as an explicit cue to communicate the environmentally friendliness of the product).
When it concerns implicit cues like packaging materials, there have been multiple innovations when it comes to "greening" packaging materials over the past years. This ranges from petbottles made from plant-based materials to removing the plastic layer on the surface of milk cartons. Importantly, these packaging materials may not only be better for the environment compared to their regular counterparts, but they might also stand out more compared to the regular packages (i.e., salience) and additionally signal implicitly to the consumers that the product is more environmentally friendly (Magnier & Schoormans, 2015;Steenis et al., 2017). In fact, this implicit signaling function (of, for instance, a carboard look) is now sometimes even misused by companies to pose a packaging as environmentally friendly while in fact it is not (e.g., plastic wrappers with a carboard look), referred to as greenwashing (Koenig-Lewis et al., 2014).
Eco-labels, in contrast, are more explicit cues. Eco-labels are visuals that "provide consumers with information about the environmental quality of individual products" (Thøgersen et al., 2010(Thøgersen et al., , p. 1787). There are many types of eco-labels. Some are vague and misleading (e.g., "dolphin friendliness"), whereas others are specific and have an objective truth value (e.g., the EU organic label; European Comission, 2020). Labels in general are known to have a positive effect, as they attract consumers' attention (Eelen et al., 2015) and are thus possibly more salient (Bhatt et al., 2021). Furthermore, labels are able to provide insights into the intrinsic attributes of a product (Herbes et al., 2020), like possibly the environmental friendliness of the product (Binninger, 2017;Meijers et al., 2019;Sundaraja et al., 2021). Although there are some contradictory findings in regard to the persuasiveness of eco-labels (e.g., D'Souza et al., 2007), most published studies have found a positive effect of eco-labels on product perceptions and preferences, such as a higher perceived quality (i.e., the belief that the product offers good taste) or a higher overall satisfaction with the product (e.g., D'Souza et al., 2006;Larceneux et al., 2012;Thøgersen et al., 2010).

Stimulus-Driven processing: Environmentally friendly cues
To assess the communicative impact of the implicit and explicit cues, a distinction is made between two types of processes, being stimulus-driven and goal-directed information processing (Yantis, 2000). We focus on stimulus-driven processing in this section and elaborate on goal-directed processing in the next section. Stimulus-driven, or bottom-up processing, is triggered by information from the environment, such as the visual field (Chandon et al., 2009). With stimulus-driven processing, consumers are likely to focus on attributes that stand out (Meißner et al., 2016). Research on this topic, focusing both on print advertisements and product packaging, indicated size and color as the most notable factors determining salience and focal attention (Ampuero & Vila, 2017;Pieters & Wedel, 2004;Pieters et al., 2010;Smit et al., 2015;Underwood, 2003;Zhang et al., 2009).
Considering stimulus-driven processing, both implicit and explicit environmentally friendly cues can be seen as relevant stimuli for increasing the perceived salience of the packaging, as the frequency of cues that signal environmental friendliness on packages in a supermarket might still be lower than the frequency of other packaging cues (IRi, 2020). Scarce research suggests that implicit cues, such as innovative packaging materials that give an appeal of being biodegradable, might indeed stand out against other packaging materials through color and material (Rundh, 2009). Furthermore, eyetracking research shows that some of the most looked-at areas of a product package are labels (i.e., explicit cues; Piqueras-Fiszman et al., 2013). This implies that both implicit and explicit cues are capable of capturing attention.
Environmentally friendly packaging material and labels are not only used to influence the salience of a product; they are also used to promote perceptions of environmental friendliness (Thøgersen et al., 2010). However, the fact that a product carries an environmentally friendly cue does not automatically mean that it is actually perceived as such. This study therefore aims to test whether and how the use of implicit and explicit environmentally friendly cues is actually successful in promoting environmentally friendly product attributes. Based on centrality theory, a packaging cue as a feature can influence the overall perception of the package and thereby of the product (Gershoff & Frels, 2015). This leads to the following hypotheses: H1. Product packages with an implicit environmentally friendly cue (packaging materials suggesting environmentally friendliness -cardboard) are more often perceived as salient (H1a) and environmentally friendly (H1b) than product packages without an implicit environmentally friendly cue (i.e., packaging materials not suggesting environmentally friendliness -white or brown).
H2. Product packages with an explicit environmentally friendly cue (i.e., an eco-label) are more often perceived as salient (H2a) and environmentally friendly (H2b) than product packages without an explicit environmentally friendly cue (i.e., no eco-label). RQ1. Does adding an explicit environmentally friendly cue (eco-label) to the product package moderate the effect of an implicit environmentally friendly cue (cardboard versus white or brown) on perceptions of salience (RQ1a) and environmental friendliness (RQ1b)?

Goal-Directed processing: The moderating effect of green consumerism
In contrast with stimulus-driven processing, goal-directed processing or top-down processing is conceptualized as a volitional, focal process depending on personal characteristics and motivation (Smit et al., 2013). In the case of environmentally friendly products, green consumerism might therefore be an important consumer characteristic. A so called 'green consumer' is an "individual that wants and knows how to satisfy his or her needs in the everyday life causing as little as possible impact on the environment" (Sharma & Joshi, 2017, p. 208). Consumers with stronger environmentally friendly beliefs indeed show more environmentally responsible buying behavior (Collins et al., 2007). Green consumerism is expected to stimulate goal-directed processing and potentially make cues more likely to be perceived as salient and environmentally friendly. As these cues are able to communicate environmentally friendliness, they are likely to especially attract consumers who prioritize environmental protection (Thøgersen, 2000). Furthermore, it is easier for those high in green consumerism to actually understand the message of being environmentally friendly (Thøgersen, 2000).
Because stimulus-driven and goal-directed processes often interact, we expect that green consumers might process environmentally friendly product packaging cues differently than non-green consumers. This notion is supported by studies showing that consumers with higher levels of environmental concerns or stronger environmental values respond differently to information about product sustainability (Bamberg, 2003;Birgelen et al., 2009) and their purchase intentions are more affected by environmentally friendly product attributes (Schuitema & De Goot, 2015). We therefore expect that those scoring high on green consumerism are more susceptible to packaging cues suggesting environmental friendliness, making it more likely that they perceive these cues as salient and environmentally friendly than those low on green consumerism. This leads to the following hypotheses: H3: Green consumerism enhances the effect of an implicit environmentally friendly cue on perceived salience (H3a) and environmental friendliness (H3b), resulting in greater perceived salience and environmental friendliness for product packages with an implicit environmentally friendly cue (cardboard, as compared to white and brown materials) in green (versus non-green) consumers.
H4: Green consumerism enhances the effect of an explicit environmentally friendly cue on perceived salience (H4a) and environmental friendliness (H4b), resulting in greater perceived salience and environmental friendliness for product packages with an explicit environmentally friendly cue (ecolabel, as compared to no label) for green (versus non-green) consumers.

Rationale for exploring a mouse tracking paradigm
Perceived salience of the package, the first dependent variable, was measured using a mouse tracking paradigm. This measurement method logs the x-and y-coordinates of the computer mouse with a minimum sampling rate of 60 times per second as participants move their mouse to one of multiple response alternatives on the screen (Freeman & Ambady, 2010). This paradigm has been successfully used in various research domains in cognitive science (for an overview, see http://www.mousetracker. org/publications) and was adopted to explore its potential in food marketing.
The rationale behind this exploration is that most research on the effectiveness of packaging cues has asked consumers to self-report about their perceptions of the product. For less volatile concepts like perceived environmental friendliness (or greenness) that are better suited for introspection, this is fine and not problematic. However, such self-reports might have disadvantages when examining communication effects that are more volatile like salience (Vandeberg et al., 2015(Vandeberg et al., , 2016. First, selfreports usually provide a static outcome measure at one point in time and might thereby miss shortlived attention effects. Second, they rely on introspection, while it is long known that people might lack the insight to report about the often latent cognitive processes that mediate the effect of a stimulus on a response such as salience (Nisbett & Wilson, 1977). Thus, because salience is characterized by rapid, volatile stimulus-driven attention processes that are known to change over milliseconds (e.g., Theeuwes, 1991) which as such are unlikely to (fully) enter awareness, the mouse tracking paradigm seems optimal as it not only provides a self-reported assessment of salience, but is importantly also able to reveal fine-grained information about the real-time evolution of mental processes by measuring the motor dynamics of the hand (Freeman & Ambady, 2010;Freeman et al., 2011).

Design
An experimental laboratory study was conducted, implementing a 3 (implicit cue: cardboard vs. white vs. brown control packaging material) x 2 (explicit cue: eco-label vs. no eco-label) within-subjects design. Adopting a mouse tracking paradigm, a within-subjects design with multiple trials was necessary to distinguish signal from noise, due to high variability in mouse trajectories between (and even within) subjects.

Procedure and participants
Participants were recruited at a Dutch university and participated in return for research credits. After giving informed consent, they were first presented with the mouse tracking exercise in which they were asked to judge the salience of several product packages by moving the computer mouse to their preferred response options. Afterward they filled in a survey with self-report measurements about perceived environmental friendliness of the product and green consumerism. The study was administered in English and the dataset, syntaxes, and full questionnaire are available on OSF. 1 A G*Power analysis of repeated measures with between-within interactions showed that, with a medium effect size of 0.25, an alpha of 0.05, two groups, and two measurements, the sample size should be at least 34 to reach a power of .80. Forty-two participants were recruited. They completed a total of 1260 (42*30) mouse tracking trials.
Participants were between 18 and 27 years old (M = 20.61, SD = 1.62) and 71.4% were female (28.6% male), none of the participants was colorblind and thus could see all the colors of the packaging. Six participants reported eating a vegetarian diet (14.3%) and three reported a vegan diet (7.1%). The sample is overall international (i.e., non-Dutch) with 73.8% of the participants living in the Netherlands for five years or less at the time of the study. Among the respondents, 83.3% were familiar with the EU organic label used in this study. The percentage of participants that were familiar with the label did not differ by level of green consumerism (x 2 (1) = 1.54, p = .214).

Stimulus materials
In a pretest, multiple product categories and packages were pretested to investigate which environmentally friendly packages should be used as stimulus materials. As a result, milk packages were chosen, because of their high realism and medium likability (the latter was chosen in order to prevent any floor or ceiling effects). 2 Especially fast-moving consumer goods such as milk are interesting products to study, as they are purchased on a regular basis. Furthermore, milk as a product is not necessarily environmentally friendly, which allows to study the specific impact of packaging attributes (Poore & Nemecek, 2018). In the mouse tracking exercise, a second product (i.e., chocolate) was used as a filler stimulus to make the exposure to the milk packages less repetitive.
The implicit environmentally friendly cue was manipulated by changing the look and feel of the packaging material (cf., Granato et al., 2022) and included three levels. In the condition with the implicit cue, the package had a cardboard design to give it the appeal of a biodegradable material (Figure 1 first column). In the control conditions without the implicit cue, white and brown packages were created (column 2 and 3). A brown package was added as an extra control condition to control for the color of the package in contrast to the (white) background, so to exclude the possibility that a higher salience and environmental friendliness was attributable to the contrasting colors per se, rather than the apparent biodegradable nature of the package. This extra control condition allows us to test the hypotheses more stringently on implicit cues.
The explicit environmentally friendly cue was manipulated using an eco-label (cf., Granato et al., 2022) and included two levels. Half of the packages included the EU organic label (Figure 1, first row), while the other half of the packages did not include this label (second row). Besides for reasons of ecological validity, we selected this well-known, often used label to rule out familiarity as a potential alternative explanation to H4. Based on the assumption that most of our sample would recognize the familiar eco-label (regardless of whether people were high or low in green consumerism), we aimed to investigate whether green consumerism as a personal characteristic -rather than familiarity with the label -drives an interaction with explicit cues. This resulted in six different packages.

Perceived salience
The assessment of perceived salience was programmed using MouseTracking Software (http://www. mousetracker.org/, also see, Freeman & Ambady, 2010). Participants were instructed to select which of two horizontally aligned images of product packages they noticed first, by moving their computer mouse and clicking on the corresponding package (with the "LEFT" button indicating the left image, and the "RIGHT" button indicating the right image). 3 Upon the start of the task, participants were exposed to These left and right labels were selected to prevent any verbal cues from giving away differences between the packages. a computer screen with a "START" button on the bottom center of the screen, a "LEFT" button on the top left side and a "RIGHT" button on the top right side. When they clicked the "START" button, the button disappeared and was replaced with the pictures of two products side by side, see, Figure 2 for an example of the set-up. Participants were instructed to initiate mouse movement immediately after clicking the start button. 4 No limit was placed on the response time per trial, as research shows that no (versus a 2.5 sec) time limit does not affect mouse trajectories or effect sizes (Kieslich et al., 2020). Each possible combination of packages was presented twice (in different order) to randomize the position of the packages and create a fully balanced within-subjects design; this resulted in a total number of 30 trials (reflecting fifteen different package combinations in two orders).
Several measures were taken from the mouse tracking data. First, we assessed which package was selected in any given combination of package pairs. Across the 30 trials, each packaging type had the opportunity to be selected a maximum of ten times. Response reflects the percentage of selections for a certain package in those ten trials (on a scale of 0-100) and reflects how often each package was judged to be most salient.
Second, we assessed the ease or certainty with which any expected decisions were made. This was done with two measures that each represent a unique characteristic of the mouse trajectories leading up to the expected response, one temporal measure and one spatial measure. Response times (RT) was the temporal measure and reflects the time it took participants to make the decision, with a shorter RT indicating more certainty. X-flips is the spatial measure and reflects the number of times the participant switched direction on the x-axis of the screen (tending to the left versus right response) while making the decision, with less x-flips indicating more certainty between the two response options. For each package, RTs and x-flips were averaged across the trials in which that package was selected. That is, if a package was selected seven out of ten times, the average is based on these seven trials. If packages were selected less than two times across trials, no average was calculated, and the case was treated as missing. Outliers in RTs and x-flips were winsorized, i.e., they were corrected by replacing values that were more than two standard deviations from the sample mean (−2SD<M< +2SD) by the value that reflected this cutoff point (i.e., exactly two standard deviations below or above the mean, respectively; Dixon, 1960).

Perceived environmental friendliness
After completing the mouse tracking exercise, participants were exposed to each of the six packages a second time in randomized order. Participants were asked to evaluate to what extent they perceived the respective product package as "green," on a 7-point scale ranging from 1 (not green at all) to 7 (very green).

Green consumerism
Green consumerism was measured by eighteen items based on earlier research on environmental concern (three items, e.g., "I am concerned about the environment"; Schuhwerk & Lefkoff-Hagius, 1995), attitudes toward green products (four items, e.g., "I like green products"; Chang, 2011), green purchase behavior (four items, e.g., "I make a special effort to buy products in biodegradable packages"; Kim & Choi, 2005;Shrum et al., 1995), and subjective knowledge (seven items, e.g., "I am very confident that I buy products in packages which are environmentally safe"; Ellen, 1994;Pieniak et al., 2010; the full list of items is available on the OSF). The scale is internally consistent with a Cronbach's alpha of .85 (M = 4.47, SD = 0.69). To be able to include this variable in the subsequent analysis, we relied on a median split and grouped responses into two equally large groups (n = 21) which are high, or low in green consumerism. 5

Effects on salience
Mixed ANOVAs were conducted to compare the main effects of implicit cue (packaging material: cardboard, white, brown) and explicit cue (eco-label, no eco-label) as within-subjects factors and green consumerism as a between-subjects factor (high, low) on each of the three salience outcome variables (responses, response times, x-flips) separately.
Responses. The mixed ANOVA showed a strong and significant main effect of implicit cue on responses (F(1.19, 80) = 7.06, p = .008, η 2 = .15). 6 Pairwise comparisons showed that the cardboard package (M = 60.83, SE = 3.46) was significantly more often identified as salient than the white package (M = 42.62, SE = 3.54, p = .035) and the brown package (M = 46.55, SE = 1.24, p = .002). White and brown packages did not significantly differ (p = 1.00). These findings significantly support H1a, showing that product packages with an implicit environmentally friendly cue are more often perceived as salient than product packages without an implicit environmentally friendly cue.
The main effect of explicit cue on responses was also significant and strong (F(1, 40) = 38.07, p < .001, η 2 = .49). Pairwise comparisons showed that the packages with an eco-label (M = 60.56, SE = 1.71) were identified significantly more often as salient than the packages without label (M = 39.44, SE = 1.71, p < .001). This supports H2a, by showing that product packages with an explicit environmentally friendly cue are more often perceived as salient than product packages without an explicit environmentally friendly cue.

5
Mixed ANOVA analysis only allows for categorical independent variables. 6 Note that Levene's test indicated unequal variances for the brown no label package (F(1, 40) = 4.18, p = .047) and for the cardboard label package (F(1, 40) = 4.67, p = .037). Since group sizes were similar, the ANOVA was assumed to be sufficiently robust to the violation of the assumption of homogeneity of variance (Pallant, 2010;Stevens, 1996). Also, Mauchly's Test of Sphericity indicated that the assumption of sphericity had been violated for implicit cue (χ 2 (2) = 44.06, p < .001), and thus the Greenhouse-Geisser correction is reported.
There also was a strong and significant interaction effect of green consumerism and explicit cue on responses (F(1, 40) = 10.25, p = .003, η 2 = .20). As illustrated in Figure 3, green consumers identified packages with an eco-label significantly more often as salient than packages without label. Non-green consumers did not show any significant differences in their salience responses. This supports H4a, namely, that an explicit cue (eco-label, as compared to no label) results in greater perceived salience in green (versus non-green) consumers. All other main and interaction effects were non-significant (p's > .129), thereby rejecting H3a about an interaction between green consumerism and implicit cues (cardboard versus white and brown packages) on salience and showing no explicit*implicit cue interaction on salience (RQ1a).
In the analysis for response times and x-flips, we combined the white and brown packages (i.e., the ones with no implicit green cues) into one group of packages, since the white packaging material with no eco-label condition had too few cases to analyze separately (i.e., it was too rarely chosen as the most salient option to be able to conduct the mouse tracking analyses regarding salience response time and salience x-flips). Table 1 shows the overall means for salience responses, salience response time, salience x-flips and greenness ratings per package.
Response times. The main effect of implicit cue on response times was significant and strong (F(1, 35) = 7.72, p = .009, η 2 = .18). Pairwise comparisons showed that the response time for cardboard packages (M = 1314.04, SE = 65.55) was significantly shorter than for no cardboard packages (M = 1380.13, SE = 77.41, p = .009), supporting H1a. The main effect of explicit cue on response times was significant and moderately strong (F(1, 35) = 4.77, p = .036, η 2 = .12). Pairwise comparisons showed that the response time for packages with an eco-label (M = 1322.57, SE = 68.89) was significantly shorter than for packages without eco-label (M = 1371.60, SE = 74.24, p = .036), supporting H2a. There were no interaction effects (p's <.190), showing no support for H3a and H4a, and showing no explicit*implicit cue interaction (RQ1a). Altogether, the results indicate that both implicit (cardboard packaging materials) and explicit (eco-label) eco-cues result in more and more certain salience judgments. Across the three measures of salience, we found support for H1a and H2a, the main effects of implicit and explicit eco-cues on salience. We also found support for an interaction between an explicit (eco-label) cue and green consumerism, but only for responses and not for response times or x-flips.

Effects on perceived environmental friendliness
A mixed ANOVA was performed with implicit cue (packaging material: cardboard, white, brown) and explicit cue (eco-label, no eco-label) as within-subjects factors and green consumerism as a betweensubjects factor (high, low) on perceived environmental friendliness (scale 1-7). The analysis showed a significant and strong main effect of implicit cue on perceived environmental friendliness (F(1.54, 80) = 37.44, p < .001, η 2 = .48). 7 Pairwise comparisons showed that the cardboard package (M = 5.05, SE = 0.16) was perceived as significantly more environmentally friendly than the white package (M = 3.48, SE = 0.15, p < .001) and the brown package (M = 4.06, SE = 0.13, p < .001). The brown package was also perceived as significantly more environmentally friendly than the white package (p < .001). Furthermore, there was a significant and strong main effect of explicit cue on perceived environmental friendliness (F(1, 40) = 145.90, p < .001, η 2 = .79). The package with an eco-label Mauchly's Test of Sphericity indicated that the assumption of sphericity had been violated for implicit cues (χ 2 (2) = 13.67, p = .001), and thus the Greenhouse-Geisser correction is reported.
(M = 5.36, SE = 0.14) was perceived as significantly more environmentally friendly than the package without an eco-label (M = 3.03, SE = 0.15, p < .001). All other main and interaction effects were nonsignificant (p's > .061). These findings support H1b and H2b by showing that product packages with an implicit cue (H1b, cardboard packaging material) and an explicit cue (H2b, eco-label) are more often perceived as environmentally friendly than product packages without an implicit or explicit ecocue. We find no support for H3b and H4b concerning interactions between more environmentally friendly cues and green consumerism on perceptions of environmental friendliness, and we find no explicit*implicit eco-cue interactions on perceived environmental friendliness (RQ1b).
In sum, results showed that both an implicit packaging eco-cue (cardboard material) as well as an explicit eco-cue (an eco-label) are generally capable of making food product packages more salient (H1a, H2a). We find both implicit and explicit eco-cues to lead to higher salience responses and shorter response times. For x-flips, the picture is slightly more nuanced. While we find an explicit cue (an eco-label) to lead to fewer x-flips (compared to no eco-label), we do not find differences in x-flips for packages with different implicit (packaging material) cues. Additionally, the results show that food packages as perceived more environmentally friendly when carrying an implicit or explicit cue (H1b, H2b). These cues, however, do not interact (RQ1). Furthermore, green consumerism mostly did not moderate this effect. That is, only for the salience response, results showed that green consumers were more sensitive to product packages with (versus without) an eco-label than non-green consumers (H4a was accepted for the response measures, but not for response time and x-flips measures; H3 and H4b were rejected).

Conclusion and discussion
This study aimed to investigate two requirements for environmentally friendly food purchasing, namely the extent to which products are (1) salient and (2) perceived as environmentally friendly. We hereby tested the impact of implicit and explicit packaging cues on salience and perceptions of environmental friendliness of food products. We present four key contributions.
Firstly, we find that not only cues that explicitly provide "proof" of environmental friendliness (i.e., eco-label), but also cues that implicitly suggest environmental friendliness (i.e., cardboard material), are capable of increasing salience and perceptions of environmental friendliness in consumers. This is in line with previous research showing that consumers utilize different cues to determine whether a product is environmentally friendly (Herbes et al., 2020). We can conclude that environmental friendly packages stand out, and are perceived as such, and thus, that they might have an advantage in ending up being considered and actually chosen by consumers (Krajbich et al., 2010;Van der Laan et al., 2015. Secondly, we find that an explicit cue also influences the perceived salience and perceived environmental friendliness. Furthermore, these effects on salience and environmental friendliness are larger for explicit (i.e., label) than implicit cues (i.e., cardboard material). This finding supports that an explicit cue such as an eco-label is still the stronger cue signaling environment friendliness, possibly because it explicitly provides information about the environmental quality of a product (Binninger, 2017;Thøgersen et al., 2010). Interestingly, we find that the two cues do not interact, meaning that they are influencing salience and environmental friendliness independently of each other. Yet, from a practical perspective, we can conclude that both implicit and explicit cues can be used as a method to promote environmental friendliness.
Thirdly, to explore whether we were able to gain additional insights into salience, we explored a mouse tracking paradigm (Freeman & Ambady, 2010). In this paradigm, the response measure provided real-time insights into people's salience judgments. Responses indicated that people judge packages with explicit (eco-label) but also implicit (cardboard material) cues to be more salient than packages without such eco-cues. Moreover, the mouse tracking paradigm also allowed for measures of response certainty, with response times providing a proxy of response certainty on a temporal dimension and x-flips providing a proxy of response certainty on a spatial dimension. Response times showed that salience judgments on packages with explicit and implicit cues were faster, and thereby more driven by certainty, than judgments on packages without such cues. The x-flips measure showed that explicit cues (eco-label versus no eco-label), but not implicit cues (cardboard versus noncardboard material), provided people with more certainty.
One way to use these additional measures is to validate the self-reported judgments. We speculate that, if salience judgments were a mere result of social desirability, these additional temporal and spatial measures might not result in similar result patterns. Since our results regarding certainty (response time, x-flips) support the findings stemming from the self-reported assessment of salience, this suggests that the salience judgments are not a result of mere social desirability and thus provide stronger evidence for the results that both implicit and explicit packaging cues can enhance packaging salience. Furthermore, these additional measures give more detailed insights into how consumers reach product judgments. The findings on the spatial dimension (i.e., x-flips) show that explicit cues, thus showing an eco-label versus no eco-label, result in more certain salience judgments. That is, when an eco-label is shown, consumers are physically less drawn in the direction of the alternative package without label and more directly move their mouse to the response representing the package with an eco-label. This finding does not hold for implicit (cardboard versus non-cardboard material) cues. These findings further suggest that explicit cues may be more effective in attracting consumers' attention than implicit cues. This could be further explored in future research.
Lastly, we examined whether green consumers are more susceptible to packaging cues suggesting environmentally friendliness, making it more likely that they perceive these cues as salient and environmentally friendly than those low on green consumerism. We find that green consumers more often identified the eco-label (versus without) as salient than non-green consumers. However, this salience judgment was not reflected in faster responses or less x-flips, nor was it reflected in a direct perception of being environmentally friendly. We also do not find differences in perceptions of different package materials (implicit cues). In other words, for non-green consumers, implicit (e.g., packaging materials) and explicit (i.e., eco-labels) cues also enhance salience and perceived environmental friendliness. This suggests that the well-documented finding of non-green consumer purchasing less environmentally friendly products (Schuitema & De Goot, 2015) cannot be attributed to nongreen consumers not seeing the environmentally friendly products, or not perceiving the product as environmentally friendly. Other than perceptions, more motivated reasons might play a role, like value-congruence or a price premium. This distinction can be further addressed in future research, as such adding to green consumer knowledge.
Our study is not without limitations which can provide suggestions for future research. We conducted this study in a laboratory setting to be able to implement the mouse tracking paradigm and complement the commonly used self-report measures. This comes with the limitation that we could only include a limited number of products in the experimental setup. Future research should expand the variety of stimuli in a larger number of trials. Furthermore, we relied on a student sample that is young and highly educated. Therefore, findings cannot be generalized beyond the population examined in this study. Future research should replicate these findings with a different, more heterogeneous sample, preferably in a more ecological valid setting.
Additionally, it could be argued that the exploratory mouse tracking salience measure is inappropriate, given that the measure is goal-directed whereas saliency is stimulus-driven. Ideally, we would have indeed adopted an objective, real time, fine-grained saliency measure without any goaldirected processes involved in the measurement process (e.g., eye tracking or ERP measurements), but these are difficult to implement in online experimental environments. Mouse tracking is notably more sensitive than retrospective self-report (because it provides valid, fine-grained, real-time information) and even self-reports might provide valid measures of attention (as they often correlate with more implicit or objective measures of attention, e.g., Keith et al., 2017;Morillas-Romero et al., 2015). Therefore, we consider mouse tracking to be a promising, fine-grained, easily implementable tool for online packaging research in academia and practice.
The findings have some practical managerial and societal implications stemming from the support that implicit and explicit packaging cues lead consumers to notice environmentally friendly products and perceive them as such. Marketers as well as policy makers benefit from the insight that eco-labels and environmentally friendly packaging materials matter, which enables them to more specifically use these cues to target consumers. It could thus be very beneficial for marketers to invest in new, innovative, more environmentally friendly packaging materials. This would be better for the environment, but also it would make their products "pop" on the shelf amongst the other products, thus drawing more consumers attention. Furthermore, it would also be beneficial for marketers to try more environmentally friendly ways of producing their food products, as this makes their product eligible for official eco-labels that enhance the salience of their products. Additionally, this implementation of environmentally friendly packaging and eco-labels will lead consumers to conclude that the product is environmentally friendly, a product trait that is valued by many consumers (e.g., D'Souza et al., 2006;Larceneux et al., 2012;Thøgersen et al., 2010). Trustworthy and responsibly implemented packaging cues can thus encourage environmentally friendly behavior and can be a first step to overcoming todays' ecological challenges.

Disclosure statement
No potential conflict of interest was reported by the author(s).